 Hi, everyone. Good morning. I want to start with a really quick introduction. I'm Catherine. This is a photo of me with my lovely co-authors Dia Das and Angela Lee of Genentech and University of Chicago, respectively. I'm really excited to speak to you on behalf of these amazing people to talk about learning centered teaching for the non-traditional data science classroom. So I want to start by setting up some context. Let's say that I'm teaching a class on data visualization in R. So we want learners to write out this code, this beautiful code on the left, and then seamlessly plot our beautiful data in R. However, there's more to this process, right? Before any plotting happens, we need to actually load this data file into R. We need to then clean it, and then maybe we can start plotting and visualizing. But even if learners have familiarity with this process, there can still be seed bumps. We might encounter errors every step of the way to the point where we don't even get to the exciting part of what we're trying to do. And this happens to me a lot as an instructor. So as instructors, we have to juggle a lot of things when we're teaching. First, we need to establish our learning goals, figure out what we want our learners to accomplish in the time that we have together. And this is best informed by understanding our learners, understanding what motivates learners to learn about a specific topic, as well as knowing their prior experience and expectations about computing and data more generally, and how this might relate to their approach to learning and problem solving. Therefore, our goals and what we know about our learners have a bidirectional relationship where these learning objectives and our knowledge of our learners inform each other. And this relationship is present in all learning contexts. So today, I'm going to talk about how this works in the non-traditional classroom where this relationship is so important, but sometimes overlooked. So my objectives for today are to identify different non-traditional data science classrooms. I want to also consider how learning objectives are informed by learners in the non-traditional classroom, and then also talk about some best practices for teaching data science skills to a diverse audience of learners. So here's a really quick roadmap where each stop of the way corresponds to each of our objectives. Well, I define what a non-traditional classroom is. We'll think about our goals and how they're informed by our learners, and then we'll consider different practices from pedagogical research in a formal classroom that can be applied to the non-traditional settings that I want to talk about. So let's get started. So first, I want to start by defining what a non-traditional classroom is. What do I mean when I say this phrase? I kind of am going to cheat by starting to talk about what a traditional classroom is. So for the purposes of today, we're going to consider a non-traditional class, not a non-traditional one, to be a formal, long-term course, what you might consider the traditional semester long class that you take in college or high school. We consider this to be an environment that holds learners accountable for engaging with course material through assignments, grades, and what not other assessments. So then what does this mean for non-traditional? What do we mean by non-traditional? For the purposes of this talk, this literally just means everything else, any learning environment that isn't what I just described. So we have this opportunity to make every environment a classroom. So we have this spectrum of learning environments from the traditional course that I just described to you to how we learn through our everyday experiences, experiential learning. And there are many opportunities that land in between. For instance, it could be a structured short-term workshop or training like those offered by software carpentries. It could be an online learning material that you peruse at your own pace. It could also be less formal structures such as peer learning formats where peers or colleagues share their skills and mentor each other. An example of this might be a local meetup. And then finally, we learn so much from experience. So many data scientists, people I just consider experts and superstars, they say that they learn by doing, and this is perhaps their favorite way of learning and I can't help but agree. So why dive into the non-traditional classroom? Why embrace this spectrum and think about these different learning environments that we have? There are so many different advantages. First of all, the best way to sharpen your skill set is to teach. Teaching is a wonderful way to engage with course concepts or just material or anything in a different way. It can help build your expertise in a specific subject or skill set. Non-traditional means that we can be particularly flexible. Our teaching isn't a degree program that needs to meet accreditation requirements. We can be creative with how we disseminate knowledge. This is also just a great way to create a learning community where we can all learn from each other. Of course, with every advantage, there's some challenges. Teachers say that there's never enough time in their semester or year-long courses to do everything they want to do and this is maybe even more so the case in a non-traditional environment. We're also limited with what we can ask learners to do beyond our face-to-face interactions with them. Finally, learners can be coming from different disciplines, different experiences or assumptions about the topic that we're going to talk about and this can impact their approach to learning. So we don't always take the time to get to know these things about our learners but it's always really important for setting our learning goals and objectives. So on that note, what are our learning objectives? First, let's may have an operational definition of what learning objectives are. Learning objectives are a statement in specific and measurable terms of the intellectual skills and abilities a student will develop in a given course. This is a very formal definition, very textbook-like, but it's very applicable to a non-traditional learning context. Through learning objectives, we get to communicate to both anyone who's in an instructional role as well as a learning role in our context and that's really, really important. So our steps for formulating a learning objective, we should consider these two different questions, especially in the non-traditional context, who are our learners and what are our logistical constraints. So first, we want our objectives to be learner-centered and so learning objectives communicate the goals and purpose of our teaching context to everyone and we want to tailor our goals to the learners in our classroom, whatever form that classroom may take. So it's helpful to ask these questions about our learners. First, what are the learners' backgrounds and interests? What is their previous experience with the topic and then also why are they interested in learning? So by answering these questions, we can determine what topics that we want to prioritize to learn about during our non-traditional context. We also can think about what are the relevant use cases and examples that we can use to help others learn. So is there a data set that's particularly relevant or a use case that we can zoom in on when we work together on different data techniques that can help motivate learners? Additionally, answering these questions and determining who our learners are can impact how we communicate with others in our learning environment. Different learners have different motivations for learning the same things. So how do you explain things in the rationale for learning them to those who are super interested in it and for those who don't see an immediate application? By understanding our learners better, we can put on different hats when communicating to better understand where our learners are coming from and where other instructors are coming from as well. As an instructor, we feel limited in time and resources, especially in our non-traditional classrooms, so we do need to consider our logistical constraints when we goal-set. So as a result, our learning objectives need to be smart. And by smart, I mean specific, measurable, achievable, relevant, and transparent. By being smart, we're creating these pertinent experiences that are attainable given our time and logistical constraints. So we want our learning objectives to be particularly specific, achievable, relevant, and we want students to understand how these characteristics play into the learning environment that we're in, making it transparent. So, for example, given the time constraints in a week-long course, we might ask learners to generate a script for a scatter plot. So our learning objective might be write a script that loads a CSV file and creates a scatter plot in R. So in a week-long, we can ask them to generate new material. In contrast, if we only have two hours with our students, we might just be equipping them with the skills to get started. So maybe for a two-hour tutorial, my objective might be identify the parts of a script that create a scatter plot in R. And these are important considerations to create flexible objectives for our different learning environments and learners. So on that note, once we're ready to know, once we have our goal set and our learning objectives articulated, what do we do now? So a couple things that I suggest that hopefully any educator here can take away from this talk is establishing a feedback practice cycle and equip learners for continued learning and practice. So first of all, we want to establish this feedback practice cycle. What do I mean by this? Generally, we don't have that much time in our learning environment. So all opportunities for practice need to be maximally informative and relevant to the learning objectives. And so that means that everything, both practice and feedback, must be goal-driven. Specifically, what we practice is directed by our learning goals and objectives and in order for feedback to be constructive, it needs to be informed by our goals and provide future directions for later practice. So for example, if we're thinking back to our learning objective, identify the parts of a script that create a scatter plot in R. Instead of asking our learners to write out this whole chunk of code, I might only ask them to zoom in on specific pieces of the script that directly relates to the learning objective and modify these, change these and see what happens. So kind of making sure that it's a little bit more targeted given the time that we have and given the context that we're in. Five minutes. Awesome. Thank you. When we give informative feedback, we're emphasizing the process and not the end product. So we're discussing and highlighting steps and double checks that are useful for future practice. And we're working as partners with fellow learners. We're asking about their process for learning and problem-solving, we're understanding where they're coming from, and then this can inform the feedback that you give in turn. And that's a really useful thing to have is this great flowing conversation where we're understanding processes on both sides. Finally, we're connecting feedback with our personal learning experiences as prior and present learners. We want to show empathy by referring to our connections with the learning process. By discussing your personal train of thought or workflow when coming across an error or mistake, you're showing learners that there are speed bumps and that's normal, there are mistakes, and that's normal, and it's, there's no right answer necessarily, and that's totally okay and part of the process. So a non-traditional classroom just has more room for this collaboration and partnership where everyone's a learner and where we can all grow from each other. So finally, thinking about how to equip our learners for continued learning. When we have a learning environment, we wanted to have, we wanted to have this effect where we're encouraging everyone there to continue working on it outside and beyond your face-to-face interactions with them. So on that note, we want to introduce learners to open source resources for further learning. So the best resources are the ones that are vetted by our trusted peers. So you become a trusted resource that can give them the scoop and the pro tips in the best places to look. But also, you should model how you use these resources. Other, show others how there are different ways that you troubleshoot. Think about how you pull up certain help documentation, how you conduct a Google search when you don't know what's going on. By sharing ideas and best practices and commiserating together, we're able to build a community to keep learning and practicing with each other. So it's important to point out these resources, show how you use them, take advantage of the local resources that might be around through a meetup, through Arley's and the Carpentries, which is the way that my co-authors and I all met. And if there aren't any of these available to you, your non-traditional learning context is a really perfect platform and segue to creating one. So creating a space for people to learn together, whether that's online or in person, I think the non-traditional classroom context is a really great catalyst for that. And this is also just a great segue into Angela's talk, which is coming up next. So some takeaways that I want to impart with you before I finish is, first of all, the non-traditional classroom is this amazing opportunity to explore creative methods for learning and to build this amazing community. However, all learning communities should have goals that are centered around the learners to create these achievable and relevant ways to learn and facilitate our learning together. And then we should equip learners for future practice by building these learning communities and having them continue what the non-traditional classroom context has started. And on that note, here are all those takeaways and I just want to say thank you and you can find me on any of these platforms as well as on the Slack channel. I'm happy to talk to you in person and offline. Thank you so much. Thank you so much, Catherine. That was fantastic. Looks like we have time for at least one question. So in the Q&A, it says for feedback, do you prefer live feedback or feedback at the end of the lesson slash session? That's a really good question. I think it really depends on your context. In the vein of having good collaboration and having a partnership with the learners that are in your context, having it be a continuous loop or somewhere where it's more intermixed as opposed to just at the end I think might work a lot better. But it really depends. I think that there are places where and there are even learners who want the full picture before they ask questions and others who are itchy to ask before everything's done. So I think there's no one size fits all to that. Thank you. Thank you so much.